public inbox for gdb-patches@sourceware.org
 help / color / mirror / Atom feed
From: Tom de Vries <tdevries@suse.de>
To: gdb-patches@sourceware.org
Subject: [PATCH 6/7] [gdb/testsuite] Use unique portnum in parallel testing
Date: Mon, 15 Apr 2024 17:56:26 +0200	[thread overview]
Message-ID: <20240415155627.22108-6-tdevries@suse.de> (raw)
In-Reply-To: <20240415155627.22108-1-tdevries@suse.de>

When instrumenting get_portnum using:
...
puts "PORTNUM: $res"
...
and running:
...
$ cd build/gdb
$ make check-parallel -j2 TESTS=gdb.server/*.exp
...
we run into:
...
Running gdb.server/abspath.exp ...
PORTNUM: 2345
...
and:
...
Running gdb.server/bkpt-other-inferior.exp ...
PORTNUM: 2345
...

This is because the test-cases are run in independent runtest invocations.

Fix this by handling the parallel case in get_portnum using:
- a file $objdir/cache/portnum to keep the portnum variable, and
- a file $objdir/cache/portnum.lock to serialize access to it.

Tested on aarch64-linux.
---
 gdb/testsuite/lib/gdbserver-support.exp | 47 ++++++++++++++++++++-----
 1 file changed, 38 insertions(+), 9 deletions(-)

diff --git a/gdb/testsuite/lib/gdbserver-support.exp b/gdb/testsuite/lib/gdbserver-support.exp
index 0f97ce9c0fd..41ad5e6cbfb 100644
--- a/gdb/testsuite/lib/gdbserver-support.exp
+++ b/gdb/testsuite/lib/gdbserver-support.exp
@@ -141,18 +141,47 @@ proc get_portnum {} {
     # starting at $initial_portnum, to avoid conflicts with hung ports.
     set initial_portnum 2345
 
-    # Currently available port number.
-    gdb_persistent_global portnum
+    if { ![info exists ::GDB_PARALLEL] } {
+	# Sequential case.
 
-    # Initialize, if necessary.
-    if { ![info exists portnum] } {
-	set portnum $initial_portnum
+	# Currently available port number.
+	gdb_persistent_global portnum
+
+	# Initialize, if necessary.
+	if { ![info exists portnum] } {
+	    set portnum $initial_portnum
+	}
+
+	# Return currently available port number, and update it.
+	set res $portnum
+	incr portnum
+	return $res
+    }
+
+    # Parallel case.
+    with_lock portnum.lock {
+	# Keep portnum file alongside the lock that guards it.
+	set portnum_file [lock_dir]/portnum
+
+	if { [file exists $portnum_file] } {
+	    set fd [open $portnum_file r]
+	    set portnum [read $fd]
+	    close $fd
+
+	    set portnum [string trim $portnum]
+	} else {
+	    # Initialize.
+	    set portnum $initial_portnum
+	}
+
+	set next_portnum [expr $portnum + 1]
+
+	set fd [open $portnum_file w]
+	puts $fd $next_portnum
+	close $fd
     }
 
-    # Return currently available port number, and update it.
-    set res $portnum
-    incr portnum
-    return $res
+    return $portnum
 }
 
 # Locate the gdbserver binary.  Returns "" if gdbserver could not be found.
-- 
2.35.3


  parent reply	other threads:[~2024-04-15 15:56 UTC|newest]

Thread overview: 15+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2024-04-15 15:56 [PATCH 1/7] [gdb/testsuite] Factor out proc get_portnum Tom de Vries
2024-04-15 15:56 ` [PATCH 2/7] [gdb/testsuite] Make portnum a persistent global Tom de Vries
2024-05-03 20:20   ` Tom Tromey
2024-04-15 15:56 ` [PATCH 3/7] [gdb/testsuite] Factor out proc with_lock Tom de Vries
2024-05-03 20:22   ` Tom Tromey
2024-04-15 15:56 ` [PATCH 4/7] [gdb/testsuite] Factor out proc lock_dir Tom de Vries
2024-05-03 20:23   ` Tom Tromey
2024-04-15 15:56 ` [PATCH 5/7] [gdb/testsuite] Move gpu-parallel.lock to cache dir Tom de Vries
2024-05-03 20:25   ` Tom Tromey
2024-04-15 15:56 ` Tom de Vries [this message]
2024-05-03 20:27   ` [PATCH 6/7] [gdb/testsuite] Use unique portnum in parallel testing Tom Tromey
2024-04-15 15:56 ` [PATCH 7/7] [gdb/testsuite] Use unique portnum in parallel testing (check//% case) Tom de Vries
2024-05-03 20:30   ` Tom Tromey
2024-05-04  7:48     ` Tom de Vries
2024-05-03 20:19 ` [PATCH 1/7] [gdb/testsuite] Factor out proc get_portnum Tom Tromey

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=20240415155627.22108-6-tdevries@suse.de \
    --to=tdevries@suse.de \
    --cc=gdb-patches@sourceware.org \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for read-only IMAP folder(s) and NNTP newsgroup(s).